Regex Not Working Correctly?

impactdni

I'm trying some regex. However, it isn't matching correctly.

I'm just using the regex:

preg_match('/^(.+).*\1/', $string, $match, NULL, $i)

With $i incrementing from 0 to strlen($string). This is used to find the longest string inside $string that repeats.

However, when dealing with VERY long strings (tested 100,000 characters), it fails.

NOTHING matches.
I am 100% sure that there are matches, expecially considering if there are 100,000 characters, and its looking for at a minimum, 2 characters that match, it should find SOMETHING.

Anyone have any ideas?

Note: I have tried with different FLAGS than "NULL", others tried were: PREG_OFFSET_CAPTURE, etc...

impactdni

Update:
A LOT of my regex isn't quite working correctly..
Here is a great example:
On a long website (string returned from CURL):

preg_match('/(.).*?\1/', $string, $match)

matches right off the bat
0 => "<HTML> <"
1 => "<"

but then:

preg_match('/(.+).*?\1/', $string, $match)

Which would give the expect SAME result. Returns FALSE (no matches).
Same thing happens if I replace the + with a *...
Any ideas?

Really frustrating me =P

Edit: More Info!
If I do a

$string = substr($string, 0, 4000);

The 2nd and 3rd sets of code work fine. But with

$string = substr($string, 0, 5000)

The 2nd and 3rd code fails again...

impactdni

Looks like I might have solved my own problem (sorry for posting in 3 posts).

If I limit the first (.+) to a set number MAX (.{1,4000}) the code succeeds!
I think there is a problem with a matching section matching over a certain number of characters...
Trying to find that number right now...

Update:
Baffled as to why this number is the one...
But it looks to be 4152 for me...

Can anyone comment on this?

impactdni

And if I keep the max at 4000, it only matches once (in a loop where it matches and replaces that match). There are obviously MORE matches, not necessarily 4000 character matches, but above the minimum length of 1.

If I turn the min/max to {1,2000} it matches twice. It seems like it is taking the max number as the minimum... Any reason this would be happening?

halojoy

Hello

Thought i would show some interests in your Research.
Can it be some 'string buffer size' ?
for example 2560 bytes

Anyway, there is a Regular Expression Functions (POSIX Extended) page at php.net
with info and links to 'regex functions'
http://php.net/regex
Maybe you can find why you get those results.
Then you can tell and enlighten rest of us. 🙂

Absolute Insight
Is that your website?

impactdni

Where did you get the number 2560 bytes?
I don't see that anywhere (I do think it has to do with some buffer).

And Yes. I do run Absolute Insight =P

halojoy

256 and 1024 and 4096 are magical numbers
But you are very right, 2560 is not.
Sorry. Will not disturb no more.

Good luck, on your own!
🙂